DE102011089397B4 - Method for monitoring an adaptive network - Google Patents

Abstract

A method for monitoring an adaptive network with respective network nodes (1, 3) and network connections (9) between the network nodes (1, 3), the network nodes (1, 3) each having a transceiver (5, 7) connected to a respective Network connection (9) is coupled, the transceiver (5, 7) being designed to provide a mean error value (MSE_i) which is representative of deviations of a received signal from predetermined reference signal values, for which a respective reference error value (MSE_REF) is provided - A respective actual error value (MSE_AV) is determined in at least one predetermined operating state (BZ_V) of the adaptive network depending on the average error value (MSE_i) provided by the respective transceiver (5, 7), - depending on the respective reference error value (MSE_REF ) and the assigned actual error value (MSE_AV) it is determined whether there is an error in the respective network connection (9) that mi t is coupled to the respective transceiver (5, 7) - an error in the respective network connection (9) is detected depending on whether an absolute deviation between the respective reference error value (MSE_REF) and the assigned actual error value (MSE_AV) exceeds a predetermined threshold value (THD).

Description

The invention relates to a method for monitoring an adaptive network with respective network nodes and network connections between the network nodes, the network nodes each having a transceiver which is coupled to a respective network connection.

Modern motor vehicles have a large number of control and monitoring devices and also a large number of sensors, which for example can also include a video camera or several video cameras, and also devices for the purpose of transmitting information or entertaining the vehicle occupants, such as a radio and / or a video player. In addition, navigation devices are also regularly used in modern motor vehicles. An efficient operation of these devices requires a data transmission system with a high data transmission rate.

So far, various bus systems have often been used in motor vehicles, such as Local Interconnect Network (LIN), Controller Area Network (CAN) or FlexRay. With the increasing complexity of the control and monitoring devices and an increasing number of the above-mentioned devices, the need for networks that have an extended protocol scope increases, for example the need for Ethernet networks. Such networks, for example in accordance with the Ethernet standard, are adaptive in the sense that they can compensate for certain interfering influences which can be caused, for example, by interference radiation on the network connection between the respective network nodes. This makes them particularly interesting for use in motor vehicles.

In particular for use in motor vehicles, however, it is desirable if they work reliably, but on the other hand it can also be avoided that units are undesirably disturbed by any interference radiation that may occur.

The document DE 103 49 600 A1 discloses a method for checking line faults in a bus system which has at least two bus users which are connected to a data bus having at least two bus lines for the purpose of data communication with one another. The bus participants can assume a recessive state and a dominant state, and an internal high potential and an internal low potential are available in the bus participants. A line fault is checked by the bus participant in the dominant state. It is carried out by comparing voltage levels on the bus lines with threshold values that are related to the internal high level or the internal low level of the bus participant.

The object on which the present invention is based is to create a method for monitoring a network which reliably detects errors in network connections between network nodes.

The object is achieved by the features of the independent patent claim.

The invention is characterized by a method for monitoring an adaptive network with respective network nodes and network connections between the network nodes. The network nodes each have a transceiver which is coupled to a respective network connection. The respective transceiver is designed to provide an average error value which is representative of deviations of a received signal from predetermined reference signal values. In a reference operating state of the adaptive network, a respective reference error value is determined as a function of the mean error value provided by the respective transceiver. The reference operating state can be given, for example, when a motor vehicle in which the adaptive network is arranged is virtually completely manufactured, that is to say, for example, is at the end of the line. In this way, the respective reference error value can then simply be determined.

Furthermore, in at least one predetermined operating state of the adaptive network, a respective actual error value is determined as a function of the mean error value provided by the respective transceiver. Depending on the respective reference error value and the assigned actual error value, it is determined whether there is an error in the respective network connection that is coupled to the respective transceiver. The specified operating state and the reference operating state are different from one another.

In this way, on the one hand, the advantage of the adaptive network can be used to compensate for disturbances that are caused, for example, by interference radiation on the network connection between the network nodes, and, on the other hand, errors in the respective network connection can be detected, for example caused by a poor cable connection , for example due to aging and / or corrosion. In this way, the respective faulty component can then simply be identified and, if necessary, exchanged and / or repaired. In this way it can be avoided that, for example, due to a changed radiation behavior of the network connections caused by the respective error and / or corresponding adjustment to compensate for the error within the network, undesirably high levels of interference radiation occur in the area of the network connections, which, for example, can have repercussions on other units, such as interfering with radio reception.

According to the invention, an error in the respective network connection is recognized, depending on whether an absolute deviation between the respective reference error value and the assigned actual error value exceeds a predetermined threshold value. In this way, the error can be identified computationally simply and yet reliably.

According to an advantageous embodiment, in the at least one predetermined operating state, the error value then provided is used to adapt the respective actual error value. In this way, the actual error value can simply be filtered and so short-term outliers of the average error value provided do not necessarily lead to the detection of the error in the respective network connection, which error may then not have been correctly assigned. In this context, use is made in particular of the knowledge that errors in the respective network connection are not reflected in a short-term deteriorated mean error value, but that this deviates correspondingly from the reference error value at least on average over a longer period of time.

In this context it is particularly advantageous if the adaptation of the respective actual error value is carried out in the form of a sliding mean value formation. This is particularly easy to implement computationally and leads to a desired smoothing of the course of the actual error value, which proves to be advantageous with regard to reliable detection of an error in the respective network connection.

According to a further advantageous embodiment, the respective mean error value provided is a square mean error value.

According to a further advantageous embodiment, the mean error values are corrected by means of a predetermined filter function as a basis for determining the respective reference error value and / or the respective actual error value. In this way, a contribution can be made to prevent small fluctuations in the line quality of the network connection from possibly leading to undesirably large fluctuations in the respective mean error value. The predefined filter function can include, for example, a linear quality function that can possibly be dependent on a transceiver type or transceiver manufacturer or an installation location in the motor vehicle.

According to a further advantageous embodiment, the respective actual error value is adapted as a function of a predetermined weighting of predetermined high deflections in the mean error values or the mean error values corrected by means of the predetermined filter function.

According to a further advantageous embodiment, in the reference operating state of the adaptive network, it is checked whether the respective reference error value lies within a predetermined reference value range and, if this is not the case, an error in the network connection or the respective network node is detected. In this way it can easily be ensured that an adaptive network that is already initially defective is initially regarded as free of errors.

According to a further advantageous embodiment, the predefined operating state is a start mode correlating to an engine start of a motor vehicle in which the network is arranged. In this way, the presence of the error in the respective network connection can easily be checked in each case at sensibly predetermined intervals.

According to a further advantageous embodiment, the predefined operating state is correlated to a change in a connection status of the respective network connection from a link-up status to a link-down status or vice versa.

According to a further advantageous embodiment, the specified operating state is given in that those network nodes located at the end points of the respective network connection have a specified functional network node operating state.

According to a further advantageous embodiment, the predefined operating state depends on at least one predefined vehicle function having a predefined state external to the network. The specified vehicle function is, for example, characteristic of whether a source of particularly high electromagnetic interference in the vicinity of the respective network connection is either active or inactive. This helps to obtain a respective reproducible mean error value.
According to a further advantageous embodiment, the adaptive network is Ethernet-based. This has the advantage that the Ethernet protocol is very powerful, for example it enables data transmission of up to 100 Mbit / s. In addition, such a network can be implemented particularly cost-effectively, particularly in connection with network connections corresponding to UTSP (Unshielded Twisted Single Pair) with regard to the network connections, which can include regular inexpensive cables and plug contacts.

According to a further aspect, a method and a corresponding device are provided for monitoring the adaptive network, in which a respective reference error value ( MSE_REF ) is provided, in at least one predetermined operating state of the adaptive network depending on the average error value provided by the respective transceiver, a respective actual error value is determined, depending on the respective reference error value and the assigned actual error value, it is determined whether an error in the respective There is a network connection that is coupled to the respective transceiver.

This has the particular advantage that there is no need to learn the reference error value in the reference operating state. A check of the network connection can possibly already take place during commissioning and, if necessary, an error can be recognized at the time of commissioning, for example at the manufacturer of the vehicle. This procedure can be particularly advantageous if, in connection with the specified filter function, the corrected mean error values are designed in such a way that they only assume a narrowly limited number of values, such as in the form of traffic light values “red”, “yellow”, “green”. The respective actual error values are then assigned accordingly.

In this way, a contribution can also be made to avoid a later impaired possibility of error detection if an undetected error was present when determining the reference error value in the reference operating state.

Advantageous configurations of the further aspect correspond accordingly to the above-mentioned advantageous configurations.

• 1 ein Netzwerk mit einer Überwachungsvorrichtung und
• 2 ein Ablaufdiagramm eines Programms, das in der Überwachungsvorrichtung abgearbeitet werden kann.

Embodiments of the invention are explained in more detail below with reference to the schematic drawings. Show it:

• 1 a network with a monitoring device and
• 2 a flowchart of a program that can be processed in the monitoring device.

An adaptive network ( 1 ) is arranged, for example, in a motor vehicle. It comprises several network nodes 1 , 3 and network connections 9 between the network nodes. The network nodes 1 , 3 each have their own transceiver 5 , 7th on. The adaptive network is, for example, Ethernet-based. The transceiver 5 , 7th includes a physical interface. The physical interface is designed to provide physical access to the network connection 9 establish and appropriately modulated signals of the network connection 9 memorize, which this then transmitted and / or to receive corresponding analog signals. The physical interface also includes, for example, an adaptive equalizer, an echo and crosstalk canceller, an automatic gain control, and a cutting unit, which is also referred to as a slicer.

The adaptive equalizer includes in particular a digital filter with filter coefficients which are adapted in order to keep a bit error rate as low as possible. This takes place in particular as a function of the determined mean error values, which are determined in particular in the slicer.

The echo and crosstalk canceller is designed to cancel echo and crosstalk that are generated while transmitting and receiving at the same time. The automatic gain control is designed to adapt the gain of the signals received in each case.

The slicer is designed to be dependent on the respectively received signal, which is for example pulse-amplitude-modulated, in particular PAM5-modulated, with regard to its symbol values, for example +2, +1, 0, -1, -2 on the basis of reference signal values specified for it, which can be, for example, +1 V, +0.5 V, 0 V, -0.5 V and -1 V. The respective signal values of the received signal that are transmitted via the respective network connection give way 9 was transmitted, regularly from the respective reference signal values. The physical interface is designed to generate an average error value depending on the deviations of the received signal from the specified reference signal values MSE_i provide that is representative of deviations of the received signal from the predetermined reference signal values.

The transceiver 5 , 7th can also include a physical coding module, which is also referred to as a physical coding sublayer module. It can also include a digital signal processing module.

The network regularly comprises a large number of network nodes 1 , 3 which, for example, in a motor vehicle can have a number of approximately 40 pieces. The respective network node 1 , 3 serves to connect a respective device, which can be, for example, a control or monitoring device of the vehicle, but which can also be, for example, a sensor such as a video camera or a device that is provided for information or entertainment purposes, such as a radio . In addition, the devices can also include output units, in particular optical output units such as a monitor. The devices can be designed, for example, to control the internal combustion engine of the vehicle and / or the navigation and / or for safety functions such as ABS, ESP and the like.

The respective network connection 9 between two respective network nodes 1 , 3 is preferably designed as a so-called unshielded twisted pair, which is also referred to in particular as an unshielded twisted pair USTP. Using an Ethernet USTP network of this type, data transmission at 100 Mbit / s can be implemented.

The adaptive network is a monitoring device 11 assigned, which can also be referred to as a device for monitoring the adaptive network. The monitoring device 11 has inputs through which they can receive, in particular, those from the respective transceivers 5 , 7th of the respective network nodes 1 , 3 provided mean error values MSE_i can receive in order to process them further. In addition, other operating variables of the motor vehicle in particular are also supplied to it. The monitoring device 11 has a data and program memory and a computing unit. At least one program is stored in the data and program memory, which program can be processed in the processing unit while the monitoring device is in operation.

Based on 2 a flowchart of the program which is processed in the monitoring device is explained in more detail.

The program is started in a step S1, in which variables can be initialized if necessary.

The procedure now explained relates in particular to a network node, that is to say for example the network node 1 . With regard to the other network nodes 3 and the transceivers trained in them 5 , 7th a corresponding program is preferably processed in the monitoring device.

In one step S3 it is checked whether a reference operating state BZ_REF of the adaptive network is present. In this context, the reference operating state BZ REF is given, for example, when the motor vehicle is virtually completely manufactured and is, for example, at the end of the line. If the adaptive network is in the reference operating state BZ_REF processing is then carried out in step S5 continued, in which the respective average error value provided MSE_i in particular, it is temporarily stored in the data and program memory.

In addition, in the step S5 several respective mean error values also spaced apart in time MSE_i are provided, then cached in the program and data memory.

In one step S7 becomes dependent on the one or those in the step S5 mean error values stored in the program and data memory MSE_i a respective reference error value MSE_REF determined. For this purpose, an averaging of the in step S5 stored mean error values MSE_i if done in the step S5 several mean error values MSE_i were saved.

The respective mean error value provided is preferred MSE_i a quadratic mean error value, which depends on the deviations of the respective transceiver 5 , 7th received signal from the predetermined reference signal values.

Optionally, in step S7 checked whether the determined reference error value MSE_REF lies in a predefined reference value range which is suitably predefined and can be determined, for example, by appropriate tests or simulation. If the respective reference error value MSE_REF lies outside the specified reference value range, an error entry is then preferably made, which then has the consequence, for example, that the respective vehicle is subjected to a renewed test or to an exchange of various components of the network.

If, however, the reference error value MSE_REF is within the specified reference value range, processing is then repeated in step S3 continued.

Is when the step is being processed S3 the reference operating state BZ_REF not taken, processing is done in one step S9 continued.

In the step S9 it is checked whether a predetermined operating state of the adaptive network is present. The specified operating status BZ_V can be, for example, a start mode that is correlated, for example, with an engine start of the motor vehicle in which the adaptive network is arranged. It can also be taken in correlation to a corresponding actuation of an access control system to the vehicle and / or an actuation of an ignition key.

Is the condition of the step S9 is not fulfilled, the processing is repeated in the step, if necessary after a predetermined waiting period S3 continued.

Is the condition of the step S9 however, it is fulfilled in one step S11 at least one respective mean error value MSE_i captured by the transceiver 5 , 7th of the respective network node 1 , 3 provided. According to the procedure of the step S5 can also in the step S11 several mean error values provided one after the other MSE_i in the program and data memory of the monitoring device 11 get saved.

In a step S13, depending on the one or the in the step S11 mean error values stored in the program and data memory MSE_i a respective actual error value MSE_AV determined. Is preferred in the step S13 on during the last processing of the step S13 actual error value already determined MSE_AV depending on the respective mean error value or values MSE i, which in step S11 saved. This can for example be carried out in the form of a corresponding averaging, in particular a sliding averaging.

In a subsequent step S15 it is checked whether there is an absolute deviation between the respective reference error value MSE_REF and the actual error value MSE_AV a predetermined threshold THD exceeds. If this is the case, then in one step S17 for an error condition in the respective network connection 9 recognized. Otherwise, a non-error condition is assumed NERR in the respective network connection 9 recognized. By doing this, the error condition ERR or the non-fault condition NERR can be recognized particularly reliably, since in the reference operating state BZ REF, if the reference error value MSE_REF is within the specified reference value range, it can be assumed that the network connection 9 between the respective network nodes 1 , 3 is error-free and is therefore the appropriate reference for the respective individual adaptive network with regard to the later comparison in the step S15 represents. In addition, the explained determination of the actual error value MSE AV makes a simple contribution to ensuring that the actual error value MSE_AV is representative of errors in the respective network connection 9 if it amounts to more than the threshold THD differs from the reference error value MSE REF and is not representative of the influence of interference radiation on the network connection 9 . In this way, a respective interference radiation can be suitably compensated for by means of the adaptive network and, however, when there is actually an error in the network connection 9 this can be reliably recognized.

Such errors in the network connection are in particular an error in a respective cable of the network connection 9 and / or a plug contact. Such a fault can for example be caused by corrosion, a bad connection, for example by a loosening of a plug contact. By recognizing either the fault condition ERR or the non-fault condition NERR in the respective network connection 9 a simple elimination of the error is possible through appropriate signaling to the driver and / or a corresponding error entry in an error memory, by appropriate replacement of the respective faulty component, for example the entire respective network connection 9 . In this way, interference caused by a correspondingly changed radiated interference from the network connection can also be avoided 9 can be caused, for example with regard to radio reception avoided or at least simply eliminated.

The transceiver is designed, for example, as a gigabit Ethernet receiver.

List of reference symbols

1, 3

Network node

5.7

Transceiver

9

Network connection

11

Monitoring device

S1 - S19

steps

BZ_REF

Reference operating status

MSE_i

mean error value

MSE_REF

Reference error value

BZ_V

specified operating status

MSE_AV

Actual error value

THD

Threshold

ERR

Error status in the respective network connection

NERR

Non-error status in the respective network connection

Claims (1)

A method for monitoring an adaptive network with respective network nodes (1, 3) and network connections (9) between the network nodes (1, 3), the network nodes (1, 3) each having a transceiver (5, 7) connected to a respective Network connection (9) is coupled, wherein the transceiver (5, 7) is designed to provide a mean error value (MSE_i) which is representative of deviations of a received signal from predetermined reference signal values in which - a respective reference error value (MSE_REF) is provided, - In at least one predetermined operating state (BZ_V) of the adaptive network, depending on the mean error value (MSE_i) provided by the respective transceiver (5, 7), a respective actual error value (MSE_AV) is determined, - Depending on the respective reference error value (MSE_REF) and the assigned actual error value (MSE_AV) it is determined whether there is an error in the respective network connection (9) which is coupled to the respective transceiver (5, 7) - an error in the respective network connection (9) is detected depending on whether an absolute deviation between the respective reference error value (MSE_REF) and the assigned actual error value (MSE_AV) exceeds a predetermined threshold value (THD).

Images